Sachini Weerasekara

weerasekara.s [AT] northeastern [DOT] edu

I am a PhD student in machine learning at Northeastern University, Boston, MA, where I am fortunate to be advised by Prof. Jacqueline Isaacs and Prof. Sagar Kamarthi. I previously earned my master’s degree at Northeastern University, Boston, MA and my bachelor’s degree from the University of Moratuwa, Sri Lanka.

I have also spent time doing machine learning research, applied AI, and building data pipelines at Takeda Pharmaceuticals, Dell Technologies, Aikium Inc , and EFL Global.

My research is focused on developing machine learning methods for scientific discovery, with applications spanning human biology, biological design, and autonomous systems. To this end, I develop methods that carry even small amounts of high-value signals much further in learning systems to generalize more effectively even with limited data. Across these problems, I am interested in the full machine learning lifecycle from building data and experimentation pipelines and analyzing large-scale datasets to developing, evaluating, and deploying models. Ultimately, my goal is to develop machine learning methods that enable models to learn robust representations, generalize beyond their training distributions, and remain reliable in real-world deployment.

LinkedIn  |  Github  |  Google Scholar  |  CV

profile photo
News
Sep 2026 Our work on post-pretraining for alignment in single cell foundation models is out! Check it out here!
Sep 2026 Our work on iterative action refinement for continuous control policies is out! Check it out here!
Dec 2025 I will be at NeurIPS 2025!
July 2025 I am joining Aikium Inc as a machine learning scientist intern to work on generative model fine-tuning for drug discovery.
Apr 2025 Our work on policy learning for optimizing robotic disassembly lot sizing with reinforcement learning was accepted at Journal of Advanced Manufacturing and Processing.
May 2025 Our work on generative AI-based single-cell representation learning was presented at ACE Drug Discovery Summit.
Apr 2025 Our recent work on generative AI-based single-cell representation learning was accepted at American Association of Cancer Research Annual Meeting.
Mar 2025 Our work on generative AI-based single-cell representation learning was presented at Nextgen Omics Spatial & Data.
May 2024 I joined Takeda Pharmaceuticals as a research scientist fellow to work on machine learning for cancer biology!
May 2024 Honored to receive Takeda Fellowship from Takeda Pharmaceuticals.
Apr 2024 Our work on reinforcement learning for robotic disassembly task control was accepted at Journal of Computers & Industrial Engineering.
Apr 2024 I gave an invited talk at REMADE Circular Economy Technology Summit & Conference on our recent work on reinforcement learning for robotic disassembly control.
Show more
Research Highlights

Prototype Guided Post-pre-training for Alignment in Single Cell Foundational Models
Sachini Weerasekara, Natasha Darras, Sagar Kamarthi, Colles Price, Jacqueline Isaacs
In review in International Conference on Learning Representations 2027
paper | code
REFINEPPO: Learning Continuous Control Policies by Iterative Action Refinement
Sachini Weerasekara, Sagar Kamarthi, Jacqueline Isaacs
In review in IEEE International Conference on Robotics and Automation (ICRA) 2027
paper | code
Conditional Biases in Large Language Models: How Biased User Turns Modulate In-Context Reasoning
Sachini Weerasekara, Sagar Kamarthi, Jacqueline Isaacs
In review in The Nations of the Americas Chapter of the Association for Computational Linguistics (NAACL) 2027
paper | code
Reinforcement Learning for Robotic Disassembly Task Control
Sachini Weerasekara, Wei Li, Jacqueline Isaacs, Sagar Kamarthi
CAIE 2024Journal of Computers & Industrial Engineering
paper | code
CellClique: Dissecting Tumor Microenvironments at the Single Cell Level Using Generative AI and Spatial Transcriptomics
Sachini Weerasekara, Natasha Darras, Nicolas Fernandez, Melinda Chen, Alina Ainbinder Colles Price, Jacqueline Isaacs
Cancer Research 2025American Association for Cancer Research Annual Meeting
paper
Improvements to Robotic Disassembly Lot Sizing with Task Planning through Reinforcement Learning
Sachini Weerasekara, Wei Li, Jacqueline Isaacs, Sagar Kamarthi
JAMP 2025 Journal of Advanced Manufacturing and Processing
paper | code
Prediction of Clinical Complication Onset Using Neural Point Processes
Sachini Weerasekara, Jacqueline Isaacs, Sagar Kamarthi
arXiv
Trends in Adopting Industry 4.0 for Asset Life Cycle Management for Sustainability: A Keyword Co-occurrence Network Review and Analysis.
Sachini Weerasekara, Zhenyuan Lu, Burcu Ozek, Jacqueline Isaacs, Sagar Kamarthi
Sustainability 2022Sustainability Journal
paper
Optimization Methodology for Departure Slots at Slot Coordinated Passenger Airports Based on Processing Element Bottlenecks.
Sachini Weerasekara, Varuna Adikariwattage, Nalaka Abeyratne
WTCR 2019World Conference on Transportation Research
paper
Recent Talks
Selected Awards
Takeda Fellow 2024, Fellowship Award from Takeda Pharmaceuticals.
Northeastern College of Engineering PhD Research Expo Award 2023, Northeastern College of Engineering.
International Symposium on Sustainable Systems and Technology (ISSST) Scholar Award 2023.
Ferretti & Yamamura Award for Research Excellence 2022, Northeastern University.
Dell Technologies USA Futuristic Data Science Shout Out Award, 2022. Dell Technologies.
Students Mentored
Graduate Students Boda Sai Revanth Reddy (Northeastern University - MS Student 2025)
Mahavir Kathed (Northeastern University - MS Student 2024)
Aartee Simran Dhomeja (Northeastern University - MS Student 2023)
Undergraduate Students Hari Sreeramagiri (University of Southern California 2022),
Teaching
IE7300 Machine Learning For Engineering: Teaching Assistant, Northeastern University, 2022 Fall, 2023 Spring, 2023 Summer.
IE7275 Data Mining in Engineering: Teaching Assistant, Northeastern University, 2021 Spring, 2022 Spring.

Template credit: [1] [2].